Polyoxymethylene (POM), classified under ISO 1043-1 as POM-H for homopolymer and POM-C for copolymer grades, is a semicrystalline acetal resin in which drawing-level dimensional control is governed by crystallization shrinkage, post-molding relaxation, and conditioning moisture content. Tolerancing POM Parts: Shrinkage and Moisture in the Drawing requires a distinction between mold shrinkage, measured from cavity to part, and post-mold dimensional drift measured after a defined conditioning interval. Molded sections typically achieve a crystalline fraction of 60 % to 80 %; the corresponding density increase from melt to solid produces volumetric shrinkage that must be partitioned into cavity scale, process parameter tolerance, and drawing notation. Material designations for extrusion and injection molding fall under ISO 9988-1, while dimensional measurement of molded specimens is normally conducted on plaques produced according to ISO 294-3. The drawing is therefore a measurement contract: without a stated conditioning atmosphere and shrinkage allowance, two inspectors can obtain different dimensions on the same POM article because the material has not reached its dimensional equilibrium.

Mold shrinkage in POM is anisotropic and thickness-dependent. Unfilled POM-H grades frequently show linear mold shrinkage between 0.018 mm/mm and 0.025 mm/mm in the flow direction, while unfilled POM-C grades are commonly 0.017 mm/mm to 0.023 mm/mm. Transverse shrinkage is typically higher by 0.002 mm/mm to 0.004 mm/mm because orientation effects relax differently across the flow path. Glass-fiber-reinforced POM-C, such as a 25 % glass-filled grade, can reduce mold shrinkage to 0.004 mm/mm to 0.010 mm/mm in the flow direction while increasing anisotropy. These ranges are not material constants; they shift with melt temperature, mold temperature, packing pressure, gate freeze time, and section thickness. The drawing must therefore carry a specific grade callout, not a generic “acetal” note.

What Mechanisms Force Post-Molding Dimensional Change in POM?

Dimensional change in POM originates from three overlapping mechanisms. First, primary crystallization during cooling reduces specific volume from the melt state to the solid state. This is the dominant source of mold shrinkage and is captured by ISO 294-4 when the part is measured after 16 h to 24 h at a standard atmosphere. Second, secondary crystallization and physical aging continue after ejection because the non-equilibrium skin and core structure relax toward a more stable lamellar state. This post-mold shrinkage is smaller than primary shrinkage but can affect hole-to-hole pitch in unfilled POM if measurements are taken immediately after molding. Third, moisture absorption increases mass and can produce a small swelling strain. POM is not as hygroscopic as polyamide; equilibrium moisture content at 23 °C and 50 % RH is generally 0.20 wt% to 0.30 wt% under ISO 62. Moisture-driven dimensional change is therefore smaller than thermal shrinkage, but it is not negligible when a bearing bore or snap-fit engagement is specified at a tolerance of 0.02 mm or less.

Measurement atmosphere is specified in ISO 291 as 23 °C and 50 % RH for standard tropical temperature. POM parts should be conditioned to mass equilibrium before inspection when the tolerance is below 0.05 mm. For a 2 mm wall section, dimensional stabilization is frequently observed within 48 h in this atmosphere, though published data for the precise time constant of a specific grade in a specific part geometry is limited. Drawing notes that omit conditioning time allow post-molding shrinkage and moisture uptake to be confused with dimensional nonconformance.

Mold Shrinkage Values for POM Homopolymer and Copolymer Grades

The shrinkage envelope used in drawing tolerance analysis should be selected from the material manufacturer’s datasheet and confirmed by a capability study on the production tool. Representative ranges for injection-molding grades measured on ISO 294-4 plaques are provided below. These values are grade-dependent and must not replace a tool-specific dimensional capability study.

For a close-fitting circular seat, the drawing may specify linear size limits under ISO 14405-1 and form tolerances under ISO 1101. A practical approach is to compute the cavity dimension by subtracting the predicted shrinkage from the target part dimension, then verify the actual shrinkage with a capability run. If the processed shrinkage falls outside the predicted range by more than 0.002 mm/mm, the tool steel is usually not at fault. Process parameters, gate geometry, or conditioning time are more common sources of deviation.

Moisture conditioning for POM inspection is comparatively mild, but its effect on tolerance verification is not zero. A part measured immediately after ejection can be larger in thickness and smaller in a long flow path than the same part conditioned at 23 °C/50 % RH for 48 h. The change is dominated by post-molding relaxation rather than by water uptake, because POM absorbs only small amounts of water. The table below summarizes conditioning states commonly encountered in tolerance audits.

In drawing practice, the tolerance block should state the reference atmosphere and conditioning interval, and the dimensional inspection plan should separate achievable process capability from design clearance. For a POM gear hub specified at 8 mm H7, measurement to ISO 286-1 before conditioning can falsely reject parts that would pass after 48 h at 23 °C/50 % RH. This is especially relevant when the mold has been cooled below the lower recommended mold temperature to reduce cycle time, because a cold mold freezes the skin before the core completes crystallization.

When Cavity Pressure Drops Below the Packing Threshold in Thin-Walled POM Parts

POM processing has a comparatively narrow dimensional process window because crystallization is rapid. For unfilled POM-H, melt temperatures are typically 200 °C to 220 °C, and mold temperatures are commonly held at 90 °C to 120 °C. POM-C often processes at 190 °C to 210 °C with mold temperatures of 60 °C to 100 °C. When a tool is operated near the lower mold-temperature limit, the frozen skin restricts pack transfer into the cavity. A cavity-pressure sensor placed near the end of the flow path will then record a peak cavity pressure below the required packing threshold, and the part will shrink more than the drawing allowance predicts.

On a 1,000 kN hydraulic injection molding machine with a 25 mm diameter screw, maintaining a residual melt cushion of 3 mm to 5 mm is required to prevent non-return valve leakage from corrupting the packing phase. If the cushion falls below 2 mm, shot-to-shot transferred volume changes, and thin-walled POM parts exhibit batch-to-batch dimensional scatter. Peak cavity pressures in the range 40 MPa to 70 MPa are frequently observed in functional POM parts when the gate is adequately sized and packing is maintained for the full gate-seal time. Holding pressure on the machine side is not a substitute for cavity-pressure measurement because nozzle, runner, and gate losses differ between hot-runner and cold-runner tools.

The gate freeze time is a critical threshold. If gate seal is not achieved before packing ends, the melt may retract from the cavity or produce sink in thicker ribs. For a gate with a diameter of 1.5 mm and a wall thickness of 2 mm, the gate seal time is typically short, but the exact value depends on melt temperature and thermal diffusivity. During process capability studies, part weight stabilizes only after gate seal; weight variation greater than 0.05 % to 0.10 % of part mass often indicates that packing is being terminated too early or that the cushion is inadequate. In such cases, dimensional spread in the flow direction can exceed 0.003 mm/mm even though the mold temperature remains stable.

POM requires adequate mold temperature control to reduce differential shrinkage between the surface and core. If mold temperature varies by more than ±5 °C across the cavity, the crystallization rate changes locally and warpage appears as a form error rather than as a linear size error. Drawing features such as flatness, circularity, and coaxiality then fail even when individual linear dimensions are within limits. In thin-walled POM moldings, a dimensional capability study should therefore instrument the cavity with pressure sensors and thermocouples, not merely record barrel setpoints.

For a POM bearing insert running against a steel shaft, the drawing specifies a clearance built from bearing bore diameter, shaft tolerance, and differential thermal expansion. POM has a coefficient of linear thermal expansion of approximately 1.1 × 10^-4 K^-1 to 1.5 × 10^-4 K^-1 in the flow direction; this is much higher than steel, so a small temperature rise closes the clearance. Dimensional tolerances derived from room-temperature measurements must include the service temperature boundary. A POM-C gear measured at 23 °C will expand at 70 °C by more than 0.05 mm over a 50 mm length; published data for a specific glass-filled grade should be used in the tolerance stack. In gears, tooth thickness and total profile deviation should be verified after conditioning to ISO 291, and the drawing should state whether tolerance classes under DIN 3962 or ISO 1328-1 apply at room temperature only.

Snap-fit engagement surfaces in POM require a different tolerance logic. Yield strain under ISO 527-2 can range from 8 % to more than 20 % depending on grade and test speed; this permits generous flexural strain but does not eliminate dimensional sensitivity in the locked position. If a snap arm is short and stiff, a shrinkage error of 0.002 mm/mm translates into a measurable change in engagement force. Moisture conditioning affects POM snap-fit force less than nylon, but post-molding shrinkage can still reduce locking overlap. For such features, the drawing should control the effective engagement dimension with a profile tolerance or a direct functional gauge dimension, and the inspection plan should separate post-molding shrinkage from moisture swelling by measuring before and after 48 h conditioning.

For pump impellers and valve bodies made from POM-C, the operational boundary includes exposure to water at moderate temperature and resistance to dilute acids. POM is not recommended for continuous exposure to strong acids, oxidizing agents, or hot water above 60 °C for extended periods because hydrolysis and chain scission can reduce molecular weight and dimensional stability. Tolerance verification under ISO 291 standard atmosphere is therefore a condition of measurement, not an optional post-process inspection step. Drawings that omit this conditioning state convert a stable semicrystalline material into an apparent dimensional problem.